From e67ccf061014e2bbf4ca2596e84a8629cc3dff78 Mon Sep 17 00:00:00 2001 From: Lysandre Debut Date: Fri, 19 Apr 2024 15:08:53 +0200 Subject: [PATCH] Transformers Metadata (#30344) --- utils/update_metadata.py | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) diff --git a/utils/update_metadata.py b/utils/update_metadata.py index 0762c4c2aa..2296d2998f 100755 --- a/utils/update_metadata.py +++ b/utils/update_metadata.py @@ -282,10 +282,40 @@ def update_metadata(token: str, commit_sha: str): ) tags_dataset = Dataset.from_pandas(tags_table) + hub_frameworks_json = hf_hub_download( + repo_id="huggingface/transformers-metadata", + filename="frameworks.json", + repo_type="dataset", + token=token, + ) + with open(hub_frameworks_json) as f: + hub_frameworks_json = f.read() + + hub_pipeline_tags_json = hf_hub_download( + repo_id="huggingface/transformers-metadata", + filename="pipeline_tags.json", + repo_type="dataset", + token=token, + ) + with open(hub_pipeline_tags_json) as f: + hub_pipeline_tags_json = f.read() + with tempfile.TemporaryDirectory() as tmp_dir: frameworks_dataset.to_json(os.path.join(tmp_dir, "frameworks.json")) tags_dataset.to_json(os.path.join(tmp_dir, "pipeline_tags.json")) + with open(os.path.join(tmp_dir, "frameworks.json")) as f: + frameworks_json = f.read() + with open(os.path.join(tmp_dir, "pipeline_tags.json")) as f: + pipeline_tags_json = f.read() + + frameworks_equal = hub_frameworks_json == frameworks_json + hub_pipeline_tags_equal = hub_pipeline_tags_json == pipeline_tags_json + + if frameworks_equal and hub_pipeline_tags_equal: + print("No updates on the Hub, not pushing the metadata files.") + return + if commit_sha is not None: commit_message = ( f"Update with commit {commit_sha}\n\nSee: "